Ras-dva Is a Novel Pit-1- and Glucocorticoid-Regulated Gene in the Embryonic Anterior Pituitary Gland

Abstract

Glucocorticoids play a role in functional differentiation of pituitary somatotrophs and lactotrophs during embryogenesis. Ras-dva was identified as a gene regulated by anterior neural fold protein-1/homeobox expressed in embryonic stem cells-1, a transcription factor known to be critical in pituitary development, and has an expression profile in the chicken embryonic pituitary gland that is consistent with in vivo regulation by glucocorticoids. The objective of this study was to characterize expression and regulation of ras-dva mRNA in the developing chicken anterior pituitary. Pituitary ras-dva mRNA levels increased during embryogenesis to a maximum on embryonic day (e) 18 and then decreased and remained low or undetectable after hatch. Ras-dva expression was highly enriched in the pituitary gland on e18 relative to other tissues examined. Glucocorticoid treatment of pituitary cells from mid- and late-stage embryos rapidly increased ras-dva mRNA, suggesting it may be a direct transcriptional target of glucocorticoids. A reporter construct driven by 4 kb of the chicken ras-dva 5′-flanking region, containing six putative pituitary-specific transcription factor-1 (Pit-1) binding sites and two potential glucocorticoid receptor (GR) binding sites, was highly activated in embryonic pituitary cells and up-regulated by corticosterone. Mutagenesis of the most proximal Pit-1 site decreased promoter activity in chicken e11 pituitary cells, indicating regulation of ras-dva by Pit-1. However, mutating putative GR binding sites did not substantially reduce induction of ras-dva promoter activity by corticosterone, suggesting additional DNA elements within the 5′-flanking region are responsible for glucocorticoid regulation. We have identified ras-dva as a glucocorticoid-regulated gene that is likely expressed in cells of the Pit-1 lineage within the developing anterior pituitary gland.

Ras-dva was identified as a transcript up-regulated between mid- and late-embryogenesis in a study investigating global gene expression changes in the chicken embryonic pituitary gland occurring around differentiation of cells in the pituitary-specific transcription factor-1 (Pit-1) lineage (1). In both mammals and birds, circulating glucocorticoids increase toward the end of embryonic development (1–4) and are thought to play a critical role in functional maturation of the pituitary through initiation of GH production in somatotrophs and prolactin (PRL) production in lactotrophs (5–9). The observed increase in pituitary ras-dva mRNA between embryonic day (e) 12 and e17 occurs concurrently with, or just before, the appearance of pituitary somatotrophs and lactotrophs in the chicken, respectively (10–13), and its level within the developing pituitary gland correlates with increasing circulating corticosterone (CORT) that occurs around this time (2, 14, 15). Based on its developmental expression pattern and correlation with circulating CORT in the chick embryo, ras-dva may be regulated by and/or mediate the effects of glucocorticoids in this tissue.

Anterior neural fold protein-1 (Anf-1), also known as homeobox expressed in embryonic stem cells-1 (Hesx1), is a repressor that regulates expression of transcription factors involved in anterior embryo patterning (16). Ras-dva was originally identified in a screen aimed at discovering targets down-regulated by Xenopus Anf-1/Hesx1 in anterior neural ectoderm (17) and subsequently determined to be an essential component in the fibroblast growth factor (FGF) signaling network required for early anterior neural plate and adjacent ectoderm patterning in Xenopus laevis embryos. During embryogenesis, Anf-1/Hesx1 expression becomes restricted to the ventral diencephalon and Rathke's pouch, the pituitary primordium, and is one of the earliest markers of the anterior pituitary gland (18). In the absence of Anf-1/Hesx1, a small number of embryos lack a pituitary gland altogether, but the majority of mice exhibit hypopituitarism ranging from isolated GH deficiency to combined pituitary hormone deficiency (19).

Not only is ras-dva known to be expressed and developmentally regulated in the anterior pituitary gland (1), it was also initially discovered as a target of a transcription factor essential for normal pituitary development (17), although in an extrapituitary context. One study has reported the presence of ras-dva mRNA in X. laevis tadpole pituitary (20). No other studies examining the presence or regulation of ras-dva in the neuroendocrine system have been reported, and there have been no published reports regarding glucocorticoid regulation of this gene. Therefore, the objectives of this study were to determine both ontogenic and tissue-specific expression patterns of chicken ras-dva mRNA, to characterize glucocorticoid regulation of ras-dva mRNA in embryonic pituitary cells, to identify regions of the ras-dva gene involved in its pituitary and/or glucocorticoid regulation, and to investigate a role for ras-dva in mediating glucocorticoid effects on pituitary hormone expression.

Materials and Methods

Sequencing of chicken ras-dva and analysis of potential regulatory regions

A putative chicken ras-dva clone was identified through random sequencing of a neuroendocrine cDNA library in combination with basic local alignment search tool comparisons (21). The plasmid containing chicken ras-dva (GenBank accession no. BM492047) was purified according to a standard protocol (22), and sequencing and assembly were performed as described previously (23). Primers (Sigma-Aldrich, St. Louis, MO) used for sequencing were SP6, Sense462, and Sense1145 for forward reactions, and T7, Antisense734, and quantitative RT-PCR (qRT-PCR) reverse primer for reverse reactions (Supplemental Table 1, published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org). Analysis of 5′- and 3′-flanking regions of the chicken ras-dva gene for predicted transcription factor binding sites was performed using the Transcription Element Search System (TESS; http://www.cbil.upenn.edu/cgi-bin/tess/tess). Default search parameters were used, except stringency was increased by changing the minimum log-likelihood ratio score from 12 to 18. The chicken ras-dva cDNA sequence was compared with the chicken genome sequence using Ensemble (http://www.ensembl.org/Gallus_gallus/Info/Index).

Animals, tissue collection, and pituitary dispersions

Embryonated broiler strain chicken eggs used for the experiment were obtained from Allen's Hatchery (Seaford, DE). On e0, eggs were placed in a 37.5 C humidified incubator and removed as necessary during the 21-d incubation period or allowed to hatch. For the e10-d7 ontogeny experiment, pituitary glands (n = 4) were collected as described previously (24). In another experiment, anterior pituitary glands from d7, d21, d35, and d48 chickens were also collected (n = 4). Tissues (n = 4) and caudal and cephalic anterior pituitary lobes (n = 4) were collected from e18 chickens to examine tissue and pituitary lobe-specific distribution of ras-dva mRNA. Tissues were immediately frozen in liquid nitrogen and stored at −80 C until RNA extraction. For cell culture experiments, pituitaries were isolated and dispersed from e11 and e18 chickens as previously described (11). For each replicate trial of a given experiment, anterior pituitaries from e11 (50-60 embryos) or e18 (10 embryos) were isolated and pooled. All procedures were approved by the Institutional Animal Care and Use Committee at the University of Maryland.

Cell culture

Unless otherwise stated, hormones and chemicals were obtained from Sigma-Aldrich, and cell culture reagents were obtained from Invitrogen (Carlsbad, CA). All kits and commercial reagents were used according to the manufacturers' instructions, unless otherwise noted. Cells were maintained in a 37.5 C, 5% CO₂ atmosphere. With the exception of experiments involving transfection (see below), dispersed pituitary cells were allowed to attach overnight in poly-l-lysine-coated culture plates (Corning Life Sciences, Lowell, MA) or 35-mm dishes (BD Biosciences, San Jose, CA) in serum-free medium (DMEM/F12) supplemented with 0.1% BSA, 5 μg/ml human insulin, 100 U/ml penicillin G, and 100 μg/ml streptomycin sulfate before the addition of inhibitors or treatment. COS-7 cells were maintained in DMEM supplemented with 10% fetal bovine serum (Equitech-Bio, Inc., Kerrville, TX) in 75-cm² flasks (Corning Life Sciences). Cells were harvested at the completion of each experiment by retrypsinization, immediately frozen in liquid nitrogen, and stored at −80 C until RNA extraction.

Plasmids, transfection, analysis of promoter activity, and flow cytometry

Two fragments of the 5′-flanking region of the ras-dva gene were amplified from chicken liver genomic DNA using primers listed in Supplemental Table 1 and Phusion high-fidelity PCR master mix (Finnzymes, Inc., Woburn, MA). Reactions contained 250 ng genomic DNA and 500 nm of each primer, and PCR cycling parameters were as follows: 98 C for 10 sec; 35 cycles of 98 C for 1 sec, 60 C for 5 sec, and 72 C for 1 min; and a final extension at 72 C for 1 min. All fragments are numbered relative to the ATG start codon. A region from −1 bp to −2009 bp was amplified using 5′-2 kb forward and 5′-flank reverse primers, and the resulting fragment was cloned into an empty reporter construct (pGL3-Basic; Promega, Madison, WI) using HindIII to generate pGL3-2kb. A region from −1999 bp to −4154 bp was amplified using 5′-4 kb forward and 5′-2 kb reverse primers, and the resulting fragment was cloned into the pGL3-2kb plasmid using EcoRI to generate pGL3-4kb. Both reporter vectors were sequenced in their entirety with vector- and insert-specific primers listed in Supplemental Table 1. The QuikChange II site-directed mutagenesis kit (Agilent Technologies, Inc., Santa Clara, CA) was used to mutate the proximal putative Pit-1 site within pGL3-2kb and distal and proximal putative glucocorticoid receptor (GR) binding sites within pGL3-4kb (Table 1) using primers (Supplemental Table 1) designed with the company's QuikChange Primer Design application (http://www.agilent.com/genomics/qcpd). All plasmids were sequenced to confirm presence of mutations, and the sequence from mutated plasmids was reanalyzed using TESS to verify that mutations did not create additional putative transcription factor binding sites. Chicken Pit-1α (25), chicken GR (26), and chicken ras-dva inserted into the pCMV-Sport6.1 expression vector (Sport6.1; Invitrogen) were used in overexpression experiments. The reporter construct containing −1727 to +48 of the chicken GH gene driving firefly luciferase (pGL3-1727) has been demonstrated to be responsive to dexamethasone in a rat pituitary cell line (27) and to CORT in primary chicken embryonic pituitary cells (28).

Table 1.

Predicted Pit-1 and GR binding sites located within the 5′-flanking region of chicken ras-dva^a

Locationb	Strand	Sequence (5′–3′)3	Mutated sequence (5′–3′)c
Pit-1
−341 to −350	+	ACAAATGCAT	CCCCCTGACG
−2167 to −2176	+	AAATATTCAT
−2502 to −2514	+	ATGAATTAATCCA
−3200 to −3208	+	AAAATGTAT
−3221 to −3230	−	ATGCATAGAT
−3368 to −3377	+	AATTAATCAC
GR
−2071 to −2080	+	AGCACAGATG	CTCCACGCGG
−4071 to−4062	+	AGAACAGCTG	CTCCACGCGG

^aPit-1 and GR binding sites were predicted using the web-based tool TESS (http://www.cbil.upenn.edu/cgi-bin/tess/tess).

^bLocations of putative binding sites are relative to the ATG start codon.

^cUnderlined nucleotides were changed to create mutated binding sites in pGL3-mPit1, pGL3-mpGR, pGL3-mdGR, and pGL3-mGR.

Cells were transfected using Lipofectamine 2000 (Invitrogen) in Opti-MEM I. In experiments with ras-dva or GH promoter constructs, e11 cells were allowed to attach to the culture plate for 2 h before transfection with indicated reporter constructs together with 20 ng pRL-SV40 (Promega), a normalization plasmid constitutively expressing renilla luciferase. After 6 h, transfection medium was replaced with cell culture medium, and cells were allowed to recover for 18 h before treatment or cultured under basal conditions for indicated times. COS-7 cells were recovered from flasks by retrypsinization in the presence of 0.03% EDTA and replated (2.5 × 10⁵/well) in growth medium for 24 h, after which they were transfected with appropriate plasmids. After 6 h, the transfection medium was replaced with serum-free DMEM supplemented with 0.1% BSA. For all experiments evaluating the ras-dva or GH promoter activity, cells were lysed and firefly (reporter gene) and renilla (normalization gene) enzyme activities were determined with the dual-reporter assay system (Promega). The promoter activity in each sample was determined by dividing firefly luciferase activity by renilla luciferase activity.

In the experiment investigating a potential role for ras-dva in regulating gene expression in the embryonic anterior pituitary gland, e11 chicken cells were cotransfected for 6 h with a green fluorescent protein (GFP) expression vector (29) and either Sport6.1 or ras-dva expression vector as described previously (23). After the transfection, cells were allowed to recover for 18 h before the addition of treatment. Cells were collected by retrypsinization, washed once with ice-cold PBS (pH 7.4), resuspended in 0.25 ml PBS, and held on ice until flow cytometric detection of GFP (fluorescence detection 530/30 nm) and collection at the University of Maryland, Department of Veterinary Medicine's Flow Cytometry Core Facility after passage through a cell-strainer cap containing 35 μm nylon mesh (BD Biosciences, Franklin Lakes, NJ) to remove clumps. Sorting of the GFP-positive population (10.0 ± 0.3% of cells; n = 3) resulted in collection of approximately 75,000 GFP-positive cells per group, which were collected directly into 0.5 ml cell lysis buffer RLT of the RNeasy minikit (QIAGEN, Valencia, CA) containing no β-mercaptoethanol, vortexed several times, and held at room temperature approximately 1 h until all samples were collected. Final volumes were measured and adjusted using nuclease-free water or buffer RLT so the RLT to sample ratio was 3.5:1. β-Mercaptoethanol was added to each sample (1 μl per 100 μl RLT buffer), and total RNA was extracted.

Quantitative RT-PCR

Total RNA was isolated from cultured cells and anterior pituitary, lung, kidney, and spleen tissue with RNeasy minikits (QIAGEN) and from hypothalamus, whole brain, liver, heart, breast muscle, and stomach tissue using RNeasy midikits (QIAGEN). Heart, breast muscle, and stomach tissue were digested with proteinase K (100 μg/ml) for 20 min at 55 C after homogenization, and all RNA extractions included on-column deoxyribonuclease digestion. Samples were quantified and reverse transcription (RT) reactions were performed using 500 ng total RNA and the Anchored-dT primer (Supplemental Table 1) and diluted as previously described (24), with the following exceptions: RT reactions were conducted using random primers (Invitrogen) in the experiment investigating the effect of CORT treatment on ras-dva mRNA stability, and reactions contained 25 ng total RNA and were left undiluted for experiments investigating effects of ras-dva on gene expression in cultured cells.

Levels of chicken ras-dva, GH, PRL, proopiomelanocortin (POMC), TSH β-subunit (TSH-β), GHRH receptor (GHRH-R), type 2 somatostatin receptor (SSTR2), total Pit-1, phosphoglycerate kinase 1 (PGK1), and β-actin (ACTB) mRNA were analyzed with cycling parameters and using primers (Supplemental Table 1) designed as previously described (24). In the ontogeny experiment, ras-dva mRNA levels in pituitary glands from e10 through d48 chickens were transformed and normalized to PGK1 mRNA as described previously (24). Pituitary PGK1 mRNA levels have been shown to change in the least consistent manner over the ages examined compared with other potential normalization genes (24). In the tissue distribution experiments and all cell culture experiments except the one investigating the effect of CORT on ras-dva mRNA stability, the amount of target gene mRNA was normalized to the amount of ACTB mRNA and transformed using equations described previously (23). Data indicating the stability of ACTB mRNA expression in these experiments is provided in Supplemental Table 2.

RT-PCR and rapid amplification of cDNA ends

A pool of cDNA for each tissue collected on e18 was made from the four replicate samples, and primers (Supplemental Table 1) that amplify full-length ras-dva cDNA were used to assess tissue distribution. Reactions conducted using GoTaq Green master mix (Promega) contained 500 nm of each primer and 1 μl template (cDNA, no RT, or water, as appropriate). Cycling parameters were as follows: 95 C for 3 min; 35 cycles of 95 C for 45 sec, 52 C for 45 sec, and 72 C for 2 min; and a final extension at 72 C for 5 min.

For rapid amplification of cDNA ends (RACE) reactions, cDNA from e18 pituitary glands and no reverse transcriptase controls were amplified with two nested PCR reactions using primers (Supplemental Table 1) indicated below and GoTaq Green master mix (Promega). For 5′-RACE, dC-tailed cDNA was produced from 1 μg total RNA extracted from e18 pituitary glands using Invitrogen's 5′-RACE system for rapid amplification of cDNA with the Antisense734 primer (Supplemental Table 1). For 3′-RACE, 1 μg total RNA from e18 pituitary glands was reverse transcribed using the Anchored-dT primer (Supplemental Table 1) and SuperScript III (Invitrogen). Initial PCR reactions were purified with a spin column-100 (Sigma-Aldrich), diluted 1:100 and subsequently amplified with nested primers. Reactions for 5′-RACE contained 1 μl template and 400 nm of each primer, and the following cycling parameters were used: 95 C for 3 min; 35 cycles of 95 C for 1 min, 60 C for 1 min, and 72 C for 3 min; and a final extension at 72 C for 10 min. Reactions for 3′-RACE contained 1 μl template, one gene-specific sense primer, and 1.5 μm Anchored-dT primer and were amplified with the following parameters: 95 C for 3 min; 35 cycles of 95 C for 45 sec, 48.2 C for 45 sec, and 72 C for 2 min; and a final extension at 72 C for 7 min. Products were visualized using agarose gel (1%) electrophoresis and ethidium bromide staining.

Data analysis

Ras-dva ontogeny qRT-PCR data are expressed and were analyzed as described previously (24). Promoter activity and other qRT-PCR data were log₂-transformed before statistical analysis using SAS software (SAS Institute, Cary, NC). To examine ras-dva mRNA tissue distribution on e18, promoter activity, glucocorticoid regulation of ras-dva mRNA levels, and effects of ras-dva on pituitary mRNA expression, data were analyzed by ANOVA using the MIXED models procedure in SAS with differences between groups determined by the test of least significant difference (PDIFF). Data were analyzed using the t- test procedure (two tailed) in the experiment evaluating mRNA levels in caudal and cephalic pituitary lobes and in the experiment evaluating CORT effects on ras-dva mRNA half-life.

Results

Chicken ras-dva is highly homologous to ras-dva from other vertebrates

A homolog for chicken ras-dva (accession no. AY729886) has been identified through screening GenBank databases using basic local alignment search tool searches for X. laevis ras-dva homologs (30) but has not been verified by direct sequencing. We sequenced the insert contained in our clone in its entirety and found that chicken ras-dva (accession no. HQ317880) is 1276 bp long, and the longest open reading frame encodes a predicted protein of 208 amino acids. Our sequence contains 122 nucleotides in the 5′-untranslated region (UTR) and 527 nucleotides in the 3′-UTR. PCR products from 5′- and 3′-RACE (Fig. 1A) were of the predicted lengths (624, 389, and 242 bp for 5′-RACE reactions 1, 2A, and 2B, respectively, and 816 bp for 3-RACE reaction 2) and were sequenced to confirm that a single variant corresponding to our clone insert exists in the embryonic pituitary gland. The final 156 nucleotides in the 3′-UTR of our sequence did not align well with that of the chicken ras-dva sequence already in GenBank, but the results and sequence of our 3′-RACE product do not support the existence of alternative transcript variants that differ in their 3′-UTR, and our sequence aligns perfectly with the ras-dva transcript (Ensembl ID ENSGALT00000034869) in the chicken genome. Ras-dva homologs were identified in one mammal (opossum) and several nonmammalian vertebrates, and comparison of the predicted amino acid sequence of chicken ras-dva with that of other species indicates a high sequence identity among vertebrates (52-92%; Supplemental Table 3 and Supplemental Fig. 1).

Fig. 1.

Chicken ras-dva mRNA is developmentally regulated, highly enriched in the anterior pituitary gland during late embryogenesis, and expressed in both caudal and cephalic lobes of the avian pituitary. A, Agarose gel pictures of 5′- and 3′-RACE PCR products and no-RT control reactions conducted on e18 pituitary RNA. Primers are listed in Supplemental Table 1. For 5′-RACE, tailed cDNA was subjected to two nested PCR reactions using the 5′RACE_AAP and Antisense624 primers for reaction 1 and the 5′RACE_AUAP primer in combination with either the qRT-PCR ras-dva reverse primer or the Antisense240 for reactions 2A and 2B, respectively. For 3-RACE, cDNA was subjected to two nested PCRs containing the Anchored-dT primer in combination with the Sense135 primer for reaction 1 or the Sense462 primer for reaction 2. B, Total RNA isolated from e10 through d48 pituitaries (n = 4) was analyzed by qRT-PCR to determine expression levels of ras-dva mRNA. The graph depicts results from two experiments, one profiling mRNA expression from e10 through d7 and the other profiling expression from d7 through d 48. For both experiments (e10 through d7 and d7 through d48), the normalized level of ras-dva mRNA in each sample was divided by the mean of normalized ras-dva mRNA levels on d7, such that data are expressed relative to d7 (equal to 100%) for each experiment and could be graphed together. Values (mean ± sem) denoted with different letters are significantly different from one another (P < 0.05) ND, Not detectable. C, Levels of ras-dva mRNA in e18 pituitary (Pi), hypothalamus (Hy), whole brain (Br), lung (Lu), kidney (Ki), spleen (Sp), liver (Li), heart (He), breast muscle (Mu), and stomach (St) (n = 4) were determined by qRT-PCR. Inset depicts levels in extrapituitary tissues. The normalized transformed value for each sample was divided by the mean of the normalized transformed value for anterior pituitary tissue, such that data are expressed relative to levels in the pituitary gland on e18 (equal to 100%). Values (mean + sem) without a common letter are significantly different (P < 0.05). D, Agarose gel picture depicting RT-PCR reactions for ras-dva and ACTB from the same e18 tissues in B. Control reactions containing no reverse transcriptase (RT−) and no template (NT) were also analyzed. E, Levels of GH, PRL, POMC, TSH-β, and ras-dva mRNA in caudal and cephalic pituitary lobes (n = 3) collected from e18 chicks. The normalized transformed value for each sample was divided by the mean of the normalized transformed value for the lobe with the highest expression level for a particular gene, such that data are expressed relative to levels in the caudal or cephalic lobe (equal to 100%). Values (mean ± sem) denoted with an asterisk indicate significantly higher levels in that lobe (P < 0.05).

Ras-dva is developmentally regulated and highly enriched in the pituitary during late embryogenesis

We characterized anterior pituitary levels of ras-dva mRNA during the second half of embryogenesis and early post hatch, the period during which initiation of hormone transcription in the five pituitary cell types occurs in mammals and birds (1, 10, 12, 31, 32). We also measured pituitary ras-dva mRNA in birds between d7 and d48 (Fig. 1B). As expected (1), ras-dva mRNA levels increased between e10 and e18 (P < 0.05). Pituitary mRNA levels then decreased just before hatch, remained low through d7, decreased between d7 and d21, and were undetectable in older birds (P < 0.05). Ras-dva mRNA expression was examined in a broad array of neuroendocrine (pituitary and hypothalamus) and other tissues known to be important glucocorticoid targets on e18, the age when levels in the anterior pituitary gland were observed to be highest. Although detected everywhere, levels of ras-dva mRNA in the pituitary were greater than 1200-fold higher than those in other tissues (Fig. 1C; P < 0.05), a finding confirmed using standard RT-PCR (Fig. 1D).

The avian anterior pituitary consists of two anatomically distinct caudal and cephalic lobes. Somatotrophs reside primarily in the caudal lobe, whereas lactotrophs, thyrotrophs, and corticotrophs are mainly localized within the cephalic lobe (10, 33–38). We measured mRNA levels of GH, PRL, POMC, TSH-β, and ras-dva in caudal and cephalic pituitary gland regions dissected from e18 chicks. Hormone mRNA levels were highly enriched in appropriate lobes, and ras-dva mRNA was easily detected in both lobes but approximately 3-fold higher in the caudal lobe (Fig. 1E; P < 0.05).

Ras-dva is directly up-regulated by glucocorticoids in embryonic pituitary cells

The expression profile of ras-dva in the developing chicken pituitary is consistent with in vivo regulation by glucocorticoids because mRNA levels increased during embryogenesis and decreased just around hatch, similar to serum CORT (2, 4, 14, 15, 39). To determine whether CORT can induce ras-dva in the embryonic pituitary, when pituitary ras-dva mRNA and circulating glucocorticoid levels are both relatively low and relatively high, cells from e11 and e18 chickens were left untreated or treated with CORT (1 nm) for the indicated times. Both short- and long-term treatments were investigated to determine whether ras-dva is more likely a direct or indirect glucocorticoid target and whether ras-dva may be subject to down-regulation upon prolonged exposure. On both ages, CORT increased ras-dva mRNA at all time points (Fig. 2A; P < 0.05). Experiments investigating mechanisms of ras-dva up-regulation by CORT were conducted on e11 pituitary cells, when circulating CORT and ras-dva mRNA levels are both low because this should reflect in vivo developmental regulation of pituitary ras-dva by glucocorticoids. To determine whether this up-regulation involved GR, e11 anterior pituitary cells were cultured in the absence and presence of CORT (1 nm) for 0.5 or 6 h with and without the GR antagonist ZK98299 (1 μm). Again, CORT increased ras-dva mRNA levels at both time points, and the increase was blocked by inclusion of ZK98299 (Fig. 2B; P < 0.05).

Fig. 2.

Glucocorticoids rapidly induce ras-dva mRNA expression in the chicken anterior pituitary gland during mid- and late embryogenesis through transcriptional activation. A, e11 (n = 4) and e18 (n = 3) anterior pituitary cells (1 × 10⁶/well) were cultured in the absence or presence of CORT for the indicated times. B, Anterior pituitary cells (1 × 10⁶/well; n = 4) from e11 chicks were cultured for 0.5 h or 6 h in the absence or presence of CORT with or without ZK98299. Cells receiving ZK98299 were pretreated for 1.5 h before the addition of CORT. C, e11 anterior pituitary cells (1 × 10⁶/well; n = 3) were cultured in the absence or presence of CORT with or without CHX for 6 h. Cells receiving CHX were pretreated for 1.5 h before the addition of CORT. D, After pretreatment for 1 h with dimethylsulfoxide (DMSO) or the transcriptional inhibitor ActD, e11 chicken anterior pituitary cells (1 × 10⁶/well; n = 3) were cultured in the absence or presence of CORT and DMSO or ActD for 6 h. A–D, Total RNA extracted from cultured cells was analyzed by qRT-PCR for ras-dva mRNA levels. The normalized transformed value for each sample was divided by the mean of the normalized transformed value for basal cells receiving no CORT or inhibitors so that data for each gene are presented as fold induction over basal mRNA levels for each experiment. Values (mean ± sem) in each panel without a common letter are significantly different (P < 0.05). E, Anterior pituitary cells (1 × 10⁶/dish; n = 4) from e11 chickens were left untreated or treated with CORT for 6 h, after which time ActD was added to the culture medium and cells were cultured for the indicated times. Total RNA was extracted, and ras-dva mRNA levels were measured using qRT-PCR after 6 h of CORT treatment (left panel and time 0 h in right panel) and from total RNA extracted from cells collected at 2, 4, 8, 12, 16, 20, and 24 h after ActD addition. In the left panel, mRNA levels are expressed as fold induction relative to levels in basal cells receiving no CORT for the first 6 h of culture before the addition of ActD, calculated as described for panels A–D. In the right panel, the transformed value for each sample was divided by the transformed value at time of ActD addition (0 h) such that mRNA levels of ras-dva for each time point are expressed as a percentage of levels at time of the ActD addition (0 h) under basal and CORT-treated conditions for each replicate (set to 100% with no variance for both basal and CORT-treated conditions and equivalent to the respective levels depicted in the left panel). The half-life (t_½; h) under basal and CORT-treated conditions was calculated for each replicate by plotting log₂-transformed percentage vs. time after ActD addition (h) and calculating the slope of the resulting line with linear regression, which was then used to determine half-life using the equation: t_½ (hours) = −1/slope. Values (mean ± sem) without a common letter are statistically different (P < 0.05).

Based on the rapid induction in both mid- and late-embryogenesis and involvement of GR, we hypothesized that ras-dva is a direct transcriptional target of glucocorticoids in the embryonic pituitary. E11 cells were treated with and without CORT (1 nm) for 6 h in the absence or presence of the protein synthesis inhibitor cycloheximide (CHX; 10 μm) or actinomycin D (ActD; 5 μg/ml), an inhibitor of transcription. Levels of ras-dva mRNA were induced by CORT, both in the absence and presence of CHX (Fig. 2C; P < 0.05), and inclusion of ActD blocked induction of ras-dva mRNA by CORT (Fig. 2D; P < 0.05). Finally, to determine whether CORT enhances ras-dva mRNA stability in addition to increasing transcriptional activation of the ras-dva gene, e11 anterior pituitary cells were cultured in the absence or presence of CORT (1 nm) for 6 h before the addition of ActD (5 μg/ml) to inhibit new gene transcription. Changes in the level of ras-dva mRNA at time points after the addition of ActD (Fig. 2E) reflect mRNA decay, allowing the calculation of mRNA half-life as a measure of stability. Although ras-dva mRNA levels were stimulated by CORT treatment (P < 0.05), the half-life under basal conditions (t_½ = 3.7 ± 0.4 h) was the same as that under CORT-treated conditions (t_½ = 3.6 ± 0.6 h; P = 0.98) and, therefore, CORT does not increase ras-dva mRNA stability.

The ras-dva promoter is highly activated in embryonic pituitary cells

Expression of ras-dva mRNA is highly specific to the anterior pituitary gland during late embryogenesis, and ras-dva appears to be a direct transcriptional target of glucocorticoids in embryonic pituitary cells. Therefore, 5 kb of the 5′-flanking region and 2 kb of the 3′-flanking region of the chicken ras-dva gene were analyzed for putative Pit-1 and GR binding sites. The binding site for Pit-1 is a rather degenerate AT-rich region, Ta/ttAT/aTT/aATT/aCAT, in which the upper-case letters are more highly conserved (40). Traditional glucocorticoid-response elements are an inverted repeat of a hexanucleotide motif with a 3-bp spacer, AGAACAnnnTGTTCT (41). Within approximately 4 kb upstream of the ATG start codon, six potential Pit-1 binding sites and two putative GR binding half-sites (one copy of the hexanucleotide motif) were identified (Table 1). We cloned two fragments of the 5′-flanking region into a luciferase reporter construct (Fig. 3A). The longer 4154-bp fragment (pGL3-4kb) contains all eight putative Pit-1 and GR binding sites, and the shorter 2009-bp fragment (pGL3-2kb) contains only the most proximal predicted Pit-1 binding site. E11 anterior pituitary cells were transfected with the empty reporter pGL3-Basic, pGL3-2kb, or pGL3-4kb and left untreated or treated with CORT (100 nm) for the final 6 or 24 h of culture. Under basal conditions, pGL3-2kb was activated 40-fold and pGL3-4kb was activated 160-fold over pGL3-basic (Fig. 3B, left panel; P < 0.05). CORT treatment for 6 and 24 h increased the activity of pGL3-4kb an additional 4-fold (Fig. 3B, right panel; P < 0.05) and did not affect pGL3-2kb activity.

Fig. 3.

The ras-dva promoter is highly activated in chicken embryonic anterior pituitary cells, induced by glucocorticoids, and contains a functional Pit-1 binding site. A, Schematic depicting the two native and four mutated ras-dva promoter-driven firefly luciferase reporter constructs containing 4154 bp (pGL3–4kb) or 2009 bp (pGL3–2kb) of the chicken ras-dva 5′-flanking region. Constructs in which the distal (pGL3-mdGR), proximal (pGL3-mpGR), or both (pGL3-mGR) GR binding sites were mutated within pGL3–4kb and the proximal Pit-1 site was mutated within pGL3–2kb (pGL3-mPit1) were created. B, Anterior pituitary cells (1 × 10⁶/well; n = 3) from e11 chickens were cultured in the absence or presence of CORT for 6 or 24 h after transfection with an empty reporter vector (pGL3-Basic; Basic), pGL3–2kb, or pGL3–4kb in combination with a renilla luciferase expression construct. The left graph depicts promoter activity under basal conditions only. Inset in the right graph depicts activity for the pGL3-Basic construct. C, e11 pituitary cells (1 × 10⁶/well; n = 3) were cultured under basal conditions after transfection with pGL3-Basic, pGL3–2kb, or pGL3-mPit1 in combination with a renilla luciferase expression construct. D, e11 pituitary cells (1 × 10⁶/well n = 3) were cultured in the absence or presence of CORT for 6 h after transfection with pGL3-Basic, pGL3–4kb, pGL3-mdGR, pGL3-mpGR, or pGL3-mGR together with a renilla luciferase expression construct. B–D, Promoter activity is expressed as fold induction over basal cells transfected with pGL3-Basic. Values (mean ± sem) without a common letter are significantly different (P < 0.05). E, COS-7 cells (n = 4) were transfected with pGL3-Basic or pGL3–2kb and the empty expression vector (pCMV-Sport6.1) or an expression vector for chicken Pit-1α. F, COS-7 cells (n = 4) were transfected with pGL3-Basic or pGL3–4kb and pCMV-Sport6.1 or an expression vector for chicken GR. E and F, Promoter activity is expressed as fold induction over basal cells transfected with Sport6.1 and pGL3-Basic. Values (mean + sem) without a common letter are significantly different (P < 0.05).

Based on these observations, two approaches were taken to investigate functionality of the most proximal Pit-1 binding site within pGL-2kb in basal ras-dva promoter activity and the two predicted GR binding sites within pGL-4kb in CORT induction of promoter activity. First, site-directed mutagenesis was used to mutate critical nucleotides in the sequences to the alternate purine or pyrimidine of their complement (i.e. A ↔ C, G ↔ T; Table 1; Fig. 3A). To assess whether the most proximal Pit-1 site is necessary for ras-dva promoter activation in pituitary cells, e11 cells were transfected with pGL3-Basic, pGL3-2kb, or pGL2-mPit1 and cultured under basal conditions (Fig. 3C). Again, pGL3-2kb reporter activity was dramatically greater than the empty reporter (P < 0.05), and although pGL3-mPit1 was still activated over pGL3-Basic, it had lower activity than pGL3-2kb (P < 0.05). To assess functionality of the predicted GR binding sites, e11 pituitary cells were transfected with pGL3-Basic, pGL3-4kb, pGL3-mdGR, pGL3-mpGR, and pGL3-mGR and treated with and without CORT (100 nm) for 6 h (Fig. 3D). Although activity of pGL3-4kb was increased dramatically by CORT (P < 0.05), mutation of either putative GR binding site alone or the two in combination had no substantial impact on activation. In the second approach, reconstitution of promoter activity through expression of Pit-1 or GR in a nonpituitary cell type with very low endogenous GR levels (COS-7) (26) was evaluated. Reporter activity in COS-7 cells transfected with pGL3-2kb in either the absence of presence of Pit-1 was higher than activity in cells transfected with pGL3-Basic (Fig. 3E; P < 0.05), indicating slight activation of the promoter, even in nonpituitary cell types. Importantly, ras-dva promoter activity in COS-7 cells transfected with an expression vector for Pit-1 was 2-fold higher than in cells transfected with Sport6.1 (P < 0.05), and only in cells cotransfected with pGL3-4kb and an expression vector for GR was promoter activity induced by CORT treatment (Fig. 3F; P < 0.05).

Ras-dva overexpression does not affect pituitary hormone expression

Two experiments were conducted to evaluate whether overexpression of ras-dva in chicken embryonic pituitary cells would alter CORT-regulated expression of pituitary hormones initiated during mid- to late embryogenesis (somatotrophs and lactotrophs). In one experiment, pituitary cells transfected with an empty reporter vector (pGL3-Basic) or a reporter construct driven by the GH 5′-flanking region (pGL3-1727) in combination with Sport6.1 or an expression vector for ras-dva were cultured in the absence or presence of CORT (100 nm) for 20 h (Fig. 4A). The CORT-induced increase in GH promoter activity (P < 0.05) was not influenced by ras-dva overexpression. In the second experiment, pituitary cells were transfected with Sport6.1 or an expression vector for ras-dva (Fig. 4, B–H), and mRNA levels for ras-dva itself, GH, PRL, GHRH-R, SSTR2, total Pit-1, and TSH-β (Supplemental Fig. 2) were analyzed. As previously reported (5–9, 42) and similar to results in Fig. 2, CORT induced mRNA levels for ras-dva, GH, and PRL (P < 0.05). CORT did not affect POMC, GHRH-R, SSTR2, or TSH-β expression. Despite successful overexpression of ras-dva (Fig. 4H; P < 0.05), there was no effect on mRNA levels for any of the genes under basal or CORT-treated conditions.

Fig. 4.

Overexpression of ras-dva does not influence glucocorticoid regulation of GH or pituitary receptors for its major hypothalamic regulators. A, Anterior pituitary cells (1 × 10⁶/well; n = 4) from e11 chickens were transfected with a firefly luciferase reporter construct containing 1727 bp of the chicken GH 5′-flanking region (pGL3–1727) or an empty reporter vector (pGL3-Basic) in combination with a renilla luciferase expression construct and an empty expression vector (Sport6.1) or an expression vector for ras-dva. Promoter activity is expressed as fold induction over basal cells transfected with pGL3-Basic and Sport6.1. B–H, e11 anterior pituitary cells (3 × 10⁶/well; n = 4) were transfected with a GFP expression vector and Sport6.1 or an expression vector for ras-dva. Cells were left untreated (Basal) or treated with CORT for 20 h, and levels of GH, GHRH-R, SSTR2, POMC, PRL, Pit-1, and ras-dva mRNA were measured by qRT-PCR in GFP-positive cells collected by flow cytometric cell sorting. The normalized transformed value for each sample was divided by the mean of the normalized transformed value for basal cells receiving no CORT and transfected with Sport6.1, such that data for each gene are presented as fold induction over basal cells transfected with Sport6.1. A–H, Values (mean ± sem) without a common letter are significantly different (P < 0.05).

Discussion

This is the first report demonstrating that ras-dva is present and regulated by glucocorticoids in the anterior pituitary gland of any species. Interestingly, homologs for ras-dva were identified in several taxa of nonmammalian vertebrates but only in one mammalian species (Supplemental Table 2). In all vertebrates, the principles of pituitary gland commitment and development are very similar, although positional location of differentiated cell types can differ (10, 43). The major cell types are localized to anatomically discrete areas in a rostral-to-caudal manner in nonmammalian vertebrates (10, 33, 34, 36–38, 44, 45), whereas cell types with a similar phenotype in the mammalian pituitary are localized to distinct regions in a more ventral-to-dorsal manner (10, 43, 46). Ras-dva may participate in a rostral-to-caudal signaling gradient unique to nonmammalian vertebrates, consistent with its postulated involvement in the signaling network essential for anterior ectoderm patterning and development of head structures in X. laevis embryos (47).

In the chick embryo, serum CORT concentrations rise from subnanomolar levels on e10 to 40-50 nm on e17, decrease slightly on e20 (2, 4, 14, 15), and remain in the 10- to 30-nm range through d7 (39). The expression profile of ras-dva mRNA in the developing anterior pituitary during the latter half of embryogenesis and early posthatch development reflects these CORT levels, indicating possible in vivo regulation by circulating glucocorticoids. Exposure of e11 anterior pituitary cells to levels of CORT reflective of those found in circulation on e14 increased ras-dva mRNA, strongly suggesting that the rise in ras-dva mRNA observed between e10 and e18 is a result of increasing circulating glucocorticoids. Ras-dva mRNA was up-regulated by CORT within 30 min and was demonstrated to be sensitive to transcriptional inhibition but insensitive to an inhibitor of protein synthesis as well as to require GR, a ligand-activated transcription factor. CORT also did not affect ras-dva mRNA stability, strongly suggesting that ras-dva is a direct transcriptional glucocorticoid target.

Examination of the 5′-flanking region of chicken ras-dva revealed two putative GR binding sites within the region that is responsive to CORT treatment, although mutagenesis of these sites did not substantially affect induction of the promoter. Classically, glucocorticoid-bound GR functions as a transcription factor to regulate gene expression, either through direct DNA binding as a homodimer at glucocorticoid-response elements or in conjunction with other factors at half-sites within composite elements or through indirect association with DNA through protein-protein interactions. The GR half-sites within the ras-dva promoter do not appear to be part of a composite element, and mutation of these sites did not affect CORT induction of ras-dva promoter activity. Therefore, it is more likely that the mechanism through which GR transcriptionally activates the ras-dva gene is through tethering of the receptor to DNA via another transcription factor. It has been demonstrated that glucocorticoids can stimulate GnRH-R mRNA expression in a pituitary gonadotroph cell line in a mechanism involving recruitment of GR to the c-Jun- and c-Fos-bound activating protein-1 (AP-1) site within the GnRH-R gene regulatory region (48). A predicted AP-1 site exists within the ras-dva regulatory region adjacent to the most distal putative Pit-1 site. It is tempting to speculate that a mechanism in which CORT-bound GR is recruited to this region by AP-1 is also involved in glucocorticoid stimulation of ras-dva expression in pituitary cells and that this transcriptional activation involves interaction with Pit-1.

The ras-dva promoter was highly activated in embryonic anterior pituitary cells, regardless of treatment with glucocorticoids. In conjunction with pituitary enrichment of ras-dva mRNA on e18, the high level of ras-dva promoter activation in pituitary cells suggests that elements within the regulatory region are stimulated by factors enriched in the anterior pituitary gland, such as Pit-1. Six potential Pit-1 binding sites were identified within 4-kb of the translational start site, and one of these is located within 350 bp. Mutation of this proximal Pit-1 site suppressed ras-dva promoter activation in pituitary cells to levels observed in nonpituitary COS-7 cells, indicating that this is an important site in pituitary induction of ras-dva expression. Although overexpression of Pit-1 in COS-7 cells increased ras-dva promoter activity, it was not fully restored, implying other cell type-specific factors are also involved in pituitary ras-dva expression. The marginal CORT induction of the ras-dva promoter in COS-7 cells transfected with GR compared with the induction in primary pituitary cells further implies that additional pituitary-specific factors are involved in full promoter activation as well. Nonetheless, demonstrated functionality of the proximal Pit-1 site and the presence of additional predicted Pit-1 binding sites suggest that ras-dva is expressed in cells of the Pit-1 lineage.

On e18, ras-dva mRNA was detected within both lobes of the chicken anterior pituitary, indicating its expression is not restricted to any particular cell type. Thyrotrophs and lactotrophs are found in the cephalic lobe, whereas somatotrophs reside in the caudal lobe (10, 34, 36–38), so ras-dva may be found in multiple Pit-1-expressing cell types. Alternatively, expression of ras-dva may be restricted to one cell type within the Pit-1 lineage and also expressed in cells that are distributed throughout the gland, such as gonadotrophs or folliculostellate cells. Although overexpression of ras-dva in e11 anterior pituitary cells did not influence hormone mRNA expression in cells of the Pit-1 lineage, this does not exclude it from regulating other aspects of cell function, including hormone secretion and/or cell proliferation. Based on its expression in organs with high secretory activity in X. laevis, it has been proposed that ras-dva GTPases may play a role in regulation of cell secretion (20, 30). The primary source of pituitary growth factors are folliculostellate cells, and pituitary FGF2 is known to regulate proliferation and hormone secretion of multiple pituitary cell types (49–51). Ras-dva was identified as a probable component of the FGF signaling network during anterior ectoderm development (30) and may mediate autocrine/paracrine effects of FGF2 within the pituitary.

Ras-dva was identified as a novel Pit-1 and glucocorticoid-regulated gene in the developing anterior pituitary gland. The expression profile of pituitary ras-dva in the embryonic, early posthatch, and mature gland indicates it may play a critical role in pituitary development. Although the developmental profile and up-regulation by glucocorticoids are consistent with initiation of hormone expression in pituitary somatotrophs and lactotrophs, our results do not support a major role for ras-dva in directly regulating GH and PRL gene expression. Glucocorticoid induction of ras-dva mRNA in pituitary cells is a result of direct transcriptional activation of the ras-dva gene. The presence of several putative Pit-1 binding sites in the 5′-flanking region of chicken ras-dva, and the demonstration that at least one of these is functional, suggests that ras-dva is expressed in cells of the Pit-1 lineage.